Perception and software architecture for mobile robotics

When developing software for autonomous mobile robots, one has to inevitably tackle some kind of perception. Moreover, when dealing with agents that possess some level of reasoning for executing their actions, there is the need to model the environment and the robot internal state in a way that it represents the scenario in which the robot operates. Inserted in the ATRI group, part of the IEETA research unit at Aveiro University, this work uses two of the projects of the group as test bed, particularly in the scenario of robotic soccer with real robots. With the main objective of developing algorithms for sensor and information fusion that could be used e ectively on these teams, several state of the art approaches were studied, implemented and adapted to each of the robot types. Within the MSL RoboCup team CAMBADA, the main focus was the perception of ball and obstacles, with the creation of models capable of providing extended information so that the reasoning of the robot can be ever more e ective. To achieve it, several methodologies were analyzed, implemented, compared and improved. Concerning the ball, an analysis of ltering methodologies for stabilization of its position and estimation of its velocity was performed. Also, with the goal keeper in mind, work has been done to provide it with information of aerial balls. As for obstacles, a new de nition of the way they are perceived by the vision and the type of information provided was created, as well as a methodology for identifying which of the obstacles are team mates. Also, a tracking algorithm was developed, which ultimately assigned each of the obstacles a unique identi er. Associated with the improvement of the obstacles perception, a new algorithm of estimating reactive obstacle avoidance was created. In the context of the SPL RoboCup team Portuguese Team, besides the inevitable adaptation of many of the algorithms already developed for sensor and information fusion and considering that it was recently created, the objective was to create a sustainable software architecture that could be the base for future modular development. The software architecture created is based on a series of di erent processes and the means of communication among them. All processes were created or adapted for the new architecture and a base set of roles and behaviors was de ned during this work to achieve a base functional framework. In terms of perception, the main focus was to de ne a projection model and camera pose extraction that could provide information in metric coordinates. The second main objective was to adapt the CAMBADA localization algorithm to work on the NAO robots, considering all the limitations it presents when comparing to the MSL team, especially in terms of computational resources. A set of support tools were developed or improved in order to support the test and development in both teams. In general, the work developed during this thesis improved the performance of the teams during play and also the e ectiveness of the developers team when in development and test phases.

In search of reliable centimeter-level indoor positioning: a smartphone-based approach

This thesis describes the design and implementation of a reliable centimeter-level indoor positioning system fully compatible with a conventional smartphone. The proposed system takes advantage of the smartphone audio I/O and processing capabilities to perform acoustic ranging in the audio band using non-invasive audio signals and it has been developed having in mind applications that require high accuracy, such as augmented reality, virtual reality, gaming and audio guides. The system works in a distributed operation mode, i.e. each smartphone is able to obtain its own position using only acoustic signals. To support the positioning system, a Wireless Sensor Network (WSN) of synchronized acoustic beacons is used. To keep the infrastructure in sync we have developed an Automatic Time Synchronization and Syntonization (ATSS) protocol with a standard deviation of the sync offset error below 1.25 μs. Using an improved Time Difference of Arrival (TDoA) estimation approach (which takes advantage of the beacon signals’ periodicity) and by performing Non-Line-of-Sight (NLoS) mitigation, we were able to obtain very stable and accurate position estimates with an absolute mean error of less than 10 cm in 95% of the cases and a mean standard deviation of 2.2 cm for a position refresh period of 350 ms.

Seamless integration of heterogeneous networks in multicast environments

Nowadays, communication environments are already characterized by a myriad of competing and complementary technologies that aim to provide an ubiquitous connectivity service. Next Generation Networks need to hide this heterogeneity by providing a new abstraction level, while simultaneously be aware of the underlying technologies to deliver richer service experiences to the end-user. Moreover, the increasing interest for group-based multimedia services followed by their ever growing resource demands and network dynamics, has been boosting the research towards more scalable and exible network control approaches. The work developed in this Thesis enables such abstraction and exploits the prevailing heterogeneity in favor of a context-aware network management and adaptation. In this scope, we introduce a novel hierarchical control framework with self-management capabilities that enables the concept of Abstract Multiparty Trees (AMTs) to ease the control of multiparty content distribution throughout heterogeneous networks. A thorough evaluation of the proposed multiparty transport control framework was performed in the scope of this Thesis, assessing its bene ts in terms of network selection, delivery tree recon guration and resource savings. Moreover, we developed an analytical study to highlight the scalability of the AMT concept as well as its exibility in large scale networks and group sizes. To prove the feasibility and easy deployment characteristic of the proposed control framework, we implemented a proof-of-concept demonstrator that comprehends the main control procedures conceptually introduced. Its outcomes highlight a good performance of the multiparty content distribution tree control, including its local and global recon guration. In order to endow the AMT concept with the ability to guarantee the best service experience by the end-user, we integrate in the control framework two additional QoE enhancement approaches. The rst employs the concept of Network Coding to improve the robustness of the multiparty content delivery, aiming at mitigating the impact of possible packet losses in the end-user service perception. The second approach relies on a machine learning scheme to autonomously determine at each node the expected QoE towards a certain destination. This knowledge is then used by di erent QoE-aware network management schemes that, jointly, maximize the overall users' QoE. The performance and scalability of the control procedures developed, aided by the context and QoE-aware mechanisms, show the advantages of the AMT concept and the proposed hierarchical control strategy for the multiparty content distribution with enhanced service experience. Moreover we also prove the feasibility of the solution in a practical environment, and provide future research directions that bene t the evolved control framework and make it commercially feasible.

Channel modelling for MIMO systems

Systems equipped with multiple antennas at the transmitter and at the receiver, known as MIMO (Multiple Input Multiple Output) systems, offer higher capacities, allowing an efficient exploitation of the available spectrum and/or the employment of more demanding applications. It is well known that the radio channel is characterized by multipath propagation, a phenomenon deemed problematic and whose mitigation has been achieved through techniques such as diversity, beamforming or adaptive antennas. By exploring conveniently the spatial domain MIMO systems turn the characteristics of the multipath channel into an advantage and allow creating multiple parallel and independent virtual channels. However, the achievable benefits are constrained by the propagation channel’s characteristics, which may not always be ideal. This work focuses on the characterization of the MIMO radio channel. It begins with the presentation of the fundamental results from information theory that triggered the interest on these systems, including the discussion of some of their potential benefits and a review of the existing channel models for MIMO systems. The characterization of the MIMO channel developed in this work is based on experimental measurements of the double-directional channel. The measurement system is based on a vector network analyzer and a two-dimensional positioning platform, both controlled by a computer, allowing the measurement of the channel’s frequency response at the locations of a synthetic array. Data is then processed using the SAGE (Space-Alternating Expectation-Maximization) algorithm to obtain the parameters (delay, direction of arrival and complex amplitude) of the channel’s most relevant multipath components. Afterwards, using a clustering algorithm these data are grouped into clusters. Finally, statistical information is extracted allowing the characterization of the channel’s multipath components. The information about the multipath characteristics of the channel, induced by existing scatterers in the propagation scenario, enables the characterization of MIMO channel and thus to evaluate its performance. The method was finally validated using MIMO measurements.

Estimating fuzzy membership functions parameters by the levenberg-marquardt algorithm

In previous papers from the authors fuzzy model identification methods were discussed. The bacterial algorithm for extracting fuzzy rule base from a training set was presented. The Levenberg-Marquardt algorithm was also proposed for determining membership functions in fuzzy systems. In this paper the Levenberg-Marquardt technique is improved to optimise the membership functions in the fuzzy rules without Ruspini-partition. The class of membership functions investigated is the trapezoidal one as it is general enough and widely used. The method can be easily extended to arbitrary piecewise linear functions as well.

Genetic programming and bacterial algorithm for neural networks and fuzzy systems design

In the field of control systems it is common to use techniques based on model adaptation to carry out control for plants for which mathematical analysis may be intricate. Increasing interest in biologically inspired learning algorithms for control techniques such as Artificial Neural Networks and Fuzzy Systems is in progress. In this line, this paper gives a perspective on the quality of results given by two different biologically connected learning algorithms for the design of B-spline neural networks (BNN) and fuzzy systems (FS). One approach used is the Genetic Programming (GP) for BNN design and the other is the Bacterial Evolutionary Algorithm (BEA) applied for fuzzy rule extraction. Also, the facility to incorporate a multi-objective approach to the GP algorithm is outlined, enabling the designer to obtain models more adequate for their intended use.

Optimização de estruturas laminadas adaptativas com controlo activo

Neste trabalho apresenta-se um modelo de elementos finitos, baseados na teoria clássica das placas, desenvolvido para a análise de controlo activo em dinâmica de estruturas de tipo placa/casca integrando sensores e actuadores piezoeléctricos. O controlo é iniciado através de uma optimização prévia do núcleo laminado de modo a diminuir a amplitude da vibração. É usado um algoritmo de controlo baseado na ligação entre as lâminas piezoeléctricas sensoras e actuadoras para obter um mecanismo de controlo da resposta dinâmica da estrutura. A resolução por elementos finitos usa um elemento placa/casca triangular plano de 3 nós, e em cuja formulação se introduz um grau de liberdade referente ao potencial eléctrico, por cada camada piezoeléctrica do elemento finito. Apresentam-se os resultados obtidos em dois exemplos ilustrativos.

A hybrid training method for B-spline neural networks

Current and past research has brought up new views related to the optimization of neural networks. For a fixed structure, second order methods are seen as the most promising. From previous works we have shown how second order methods are of easy applicability to a neural network. Namely, we have proved how the Levenberg-Marquard possesses not only better convergence but how it can assure the convergence to a local minima. However, as any gradient-based method, the results obtained depend on the startup point. In this work, a reformulated Evolutionary algorithm - the Bacterial Programming for Levenberg-Marquardt is proposed, as an heuristic which can be used to determine the most suitable starting points, therefore achieving, in most cases, the global optimum.

Controlo activo de ressonância em estruturas piezolaminadas

Neste trabalho apresenta-se um modelo de elementos finitos baseado na teoria de deformação de corte de 3ª ordem, o qual é aplicado ao controlo activo de vibrações, incluindo o fenómeno de ressonância, em estruturas laminadas. Sensores e actuadores piezoeléctricos na forma de lâminas estão colocadas na superfície superior e inferior do laminado, permitindo assim um sistema de controlo, ligando os efeitos piézoeléctricos directo e converso, atrvés de um algoritmo baseado na realimentação com velocidade negativa. As estruturas são forçadas a vibrar num determinado modo, e a sua amplitude no tempo é calculada usando o método de Newmark. Apresenta-se uma aplicação ilustrativa.

Calibração e validação do modelo espectral de previsão da agitação marítima SWAN em zonas costeiras

Dissertação mest., Oceanografia, Universidade do Algarve, 2009

Spectral analysis of embolic signals

Tese de dout., Engenharia Electrónica e Computação, Faculdade de Ciências e Tecnologia, Univ. do Algarve, 2005

Paralelização automática de algoritmos matriciais

Tese de dout., Engenharia Electrónica e Computação, Faculdade de Ciências e Tecnologia, Univ. do Algarve, 2003

Utilização de redes TIN em modelos distribuídos de precipitação/escoamento superficial

A comunicação apresenta as directivas de desenvolvimento de um modelo distribuído e a sua aplicação à Ribeira da Pradiela (distrito de Évora, afluente do Dejebe). A utilização de modelos distribuídos de precipitação/escoamento superficial para modelação integrada dos fenómenos hidrodinâmicos, transporte de sedimentos e qualidade da água à escala da bacia hidrográfica, é fundamental para uma correcta percepção dos binómios causas/efeitos e consequente gestão optimizada dos recursos hídricos. Os modelos distribuídos existentes assentam em malhas estruturadas de células regulares adjacentes (DEM) ou em malhas de triângulos irregulares adjacentes (TIN). Os primeiros têm um processamento relativamente simples e são compatíveis com a estrutura matricial de uma imagem raster, contudo obrigam a uma densidade de informação uniforme sobre a área em estudo, apresentam problemas de escala na representação das linhas de água e das variáveis distribuídas como a altitude, classes de solo, classes de uso do solo, dotações de rega, fertilizantes e pesticidas. As redes TIN são mais versáteis e no caso particular de os pontos levantados sobre o terreno formarem uma métrica regular, a rede TIN representa uma malha estruturada. Os modelos existentes que utilizam uma malha TIN definem a rede hidrográfica pelas arestas comuns dos triângulos que formam um ângulo côncavo entre si. Esta implementação pode forçar a existência de linhas de água onde não existem traços morfológicos da sua existência. A metodologia proposta para gerar as linhas de água passa por definir uma área de influência associada a cada nó (polígono de Voronoi) que drena por uma linha de água definida entre o próprio nó e o nó vizinho com o qual forma o maior declive. Esta metodologia é relativamente simples, não cria ambiguidades e permite criar uma rede hidrográfica cuja estrutura é perfeitamente compatível com a implementação de um algoritmo de diferenças finitas. As secções transversais das linhas de água são definidas como função da área a montante da respectiva secção. A precipitação efectiva é calculada à escala da bacia pelo método da curva número (CN) do Soil Conservation Service (SCS) e à escala de uma parcela pela equação de Richardson. O escoamento é resolvido pela equação da onda cinemática nas linhas de água de ordem superior, pela equação de conservação da massa nas depressões e pelas equações de Saint-Venant na linha de água principal. É considerado o destacamento nas encostas e o destacamento/ transporte/ deposição nas linhas de água. Também é considerado o transporte de poluentes adsorvidos aos sedimentos, bem como fenómenos de wash-off .

Deteção automática de eventos em sinais Doppler de fluxo sanguíneo

Os sinais de fluxo sanguíneo são sinais aleatórios que, além de variarem de individuo para individuo, variam também consoante o ciclo cardíaco em análise. Sendo o diagnóstico de patologias cardiovasculares fortemente baseado em sinais de ultrassom Doppler representados em forma de espectrograma, e tendo em consideração que o ruído do tipo speckle é parte integrante dos sinais ultrassónicos, torna-se vital a pesquisa de métodos de eliminação desse tipo de ruído e de caracterização precisa dos parâmetros desses sinais biomédicos por forma a melhorar a qualidade do diagnóstico clínico. Com esta tese pretende-se desenvolver uma ferramenta computacional que possibilite a extração automática dos parâmetros pico sistólico, fim de diástole e de outros eventos clinicamente relevantes de sinais Doppler de fluxo sanguíneo, com especial atenção ao sinal proveniente da artéria carótida. Esta investigação vem na continuidade de um projeto realizado anteriormente no Grupo de Processamento de Sinal Biomédico da Universidade do Algarve, no qual foi desenvolvido um sistema de redução de ruído de espectrogramas Doppler. Este sistema de remoção de ruído será aqui aplicado e melhorado. Para a deteção e extração automática de parâmetros clínicos, foi desenvolvido um algoritmo que recebe um sinal de Doppler e que tem como saída o espectrograma livre de ruído e os valores dos parâmetros clínicos calculados. O algoritmo desenvolvido está dividido em três partes principais. A primeira, consiste na transformada do sinal para os domínios tempo-frequência para a criação do espectrograma e na aplicação de uma metodologia de remoção do ruído do tipo speckle do espectrograma. A fase seguinte é a criação de um sinal bidimensional a partir do espectrograma, o qual é criado para possibilitar a sua caraterização. Por último, desenvolveu-se uma fase dedicada à caracterização do sinal, tendo como principais funções a deteção dos eventos clínicos de pico sistólico, fim de diástole, índice de pulsação, índice de resistência e ratio sístole-diástole. A refinação de atuação de cada uma das três partes mencionadas pode ser ajustada pelo utilizador, tendo para o efeito sido desenvolvido uma interface gráfica na qual a interação do utilizador com o programa global é facilitada. Versatilidade e eficácia do algoritmo desenvolvido foram demonstradas pelos resultados obtidos com três sinais de Doppler de diferentes origens: um de origem clínica, um sinal de fluxo em artéria carótida simulado experimentalmente recorrendo a phantoms, e um sinal de fluxo simulado computacionalmente. Para cada um destes sinais são apresentados os valores das variáveis considerados preferenciais para o ajuste ao respetivo sinal e os respetivos espetros de ruído reduzido. Os resultados da extração automática dos parâmetros clínicos dos três sinais, comprovou a utilidade clínica do algoritmo desenvolvido.

Design of neuro-fuzzy models by evolutionary and gradient-based algorithms

All systems found in nature exhibit, with different degrees, a nonlinear behavior. To emulate this behavior, classical systems identification techniques use, typically, linear models, for mathematical simplicity. Models inspired by biological principles (artificial neural networks) and linguistically motivated (fuzzy systems), due to their universal approximation property, are becoming alternatives to classical mathematical models. In systems identification, the design of this type of models is an iterative process, requiring, among other steps, the need to identify the model structure, as well as the estimation of the model parameters. This thesis addresses the applicability of gradient-basis algorithms for the parameter estimation phase, and the use of evolutionary algorithms for model structure selection, for the design of neuro-fuzzy systems, i.e., models that offer the transparency property found in fuzzy systems, but use, for their design, algorithms introduced in the context of neural networks. A new methodology, based on the minimization of the integral of the error, and exploiting the parameter separability property typically found in neuro-fuzzy systems, is proposed for parameter estimation. A recent evolutionary technique (bacterial algorithms), based on the natural phenomenon of microbial evolution, is combined with genetic programming, and the resulting algorithm, bacterial programming, advocated for structure determination. Different versions of this evolutionary technique are combined with gradient-based algorithms, solving problems found in fuzzy and neuro-fuzzy design, namely incorporation of a-priori knowledge, gradient algorithms initialization and model complexity reduction.